1. Causes, pathophysiology and clinical manifestations of CAFs
CAFs were either congenital or acquired (1). More than 90% of CAFs were congenital or sporadic (2). The embryological mechanism of congenital CAF in the heart cavity was the degeneration failure of the peritrabecular sinus system in the myocardium (3). Acquired CAFs were often iatrogenic complications (4–7) of deceleration accidents, percutaneous transluminal coronary angioplasty, endomyocardial biopsy, implantation of permanent ventricular pacing leads, coronary artery bypass graft surgery or stent placement, mitral valve replacement, septal muscle resection or acute myocardial infarction, trauma, chest irradiation and etc. Among 62 CAFs, a 59-year-old male patient with coronary heart disease who underwent coronary stent implantation in 2012 was found no CAFs during CT follow-up in 2015, while he was diagnosed of left circumflex coronary artery fistula to right ventricle by DSCT when re-admitted to hospital due to chest pain in 2017. It was suggested that the CAFs of this patient was secondary.
The basic pathological mechanism of CAFs was that high-pressure blood flow in normal coronary arteries bypassed myocardial arterioles and capillaries, entered low-pressure heart cavities or cardiopulmonary vessels through fistula vessels, and caused myocardial stealing. Myocardial stealing made these patients more prone to symptoms of cardiothoracic discomfort than normal people under conditions of increased oxygen demand such as doing slightly more intense sports, becoming pregnancy and other special physiological conditions. Among 62 CAFs, a 25-year-old female patient suffered the first attack of palpitation with no obvious cause during pregnancy and suffered palpitation accompanied by lip cyanosis again three years later as diagnosed of left artery fistula to right ventricle. It was believed that her increased oxygen demand during pregnancy made the symptoms of coronary artery fistula appear earlier. In fact, nearly half of the clinical cases of CAFs were asymptomatic, and those with symptoms often came to the clinic with chief complaints such as chest pain, chest tightness, and palpitations and etc. It was suggested that the symptoms or severity of CAFs depended not only on the drainage site of the fistula (type of fistula), the diameter of the fistula vessels, the size of the fistula (the amount of shunts) and etc., but also on the patient's age, physical tolerance, complication of other cardiopulmonary diseases and the duration of the diseases. For instance, coronary artery-cardiac fistula was usually a single bulky fistula vessel with a large vascular shunt, and continuous murmurs during diastole and systole in clinic can be easily heard on a patient with this fistula, so patients often came to the clinic at the early stage. Among 62 cases of CAFs there were 19 cases of coronary artery-cardiac fistula including 12 patients (63%) under 30 years old. While coronary artery-pulmonary fistula, due to its small fistula and small shun, was generally asymptomatic or became symptomatic late. For example, among 62 CAFs, there were 31 cases of Coronary pulmonary-artery fistula including 4 patients under 50 years of age and 27 patients (87%) over 50 years old. Other patients came to the clinic with symptoms of myocardial theft, such as arrhythmia, angina pectoris, and myocardial infarction (8–12). Complications of congestive heart failure, endocarditis, thrombosis, pulmonary hypertension, or embolic events may not occur until late in the course of the disease.
2. Comparison of image inspection methods of CAFs
Traditional Coronary Angiography (cag)
Before the emergence or widespread use of cardiac CT, CAG was considered a diagnostic reference for CAFs. CAG itself was also one of the treatments for CAFs. Generally, CAG can well show the origin of the fistula vessels and the proximal movement, but sometimes cannot show clearly the distal fistula vessels and orificium fistulae due to the dilution of the contrast agent. In addition, CAG only provided two-dimensional projection images, making it limited to diagnose CAFs with complex drainage sites and small orificium fistulae. It was reported that only 35% -50% of CAFs were correctly diagnosed by CAG (13, 14).
Transthoracic Or Transesophageal Echocardiography
ECHO has no ionizing radiation. Although the use of microbubbles can help determine more precisely the location and extent of CAFs, ECHO has poor evaluation of CAFs with multiple fistula vessels, complex drainage sites and relatively small orificium fistulae. For example, as the left ventricular systolic pressure was the same as the aortic pressure, color Doppler was not sensitive to the abnormal flow of the fistula-entered pulmonary artery or the left ventricle, reducing the detection rate of coronary artery-pulmonary fistula. Furthermore, ECHO often depended on the operator's experience and technique.
Mr Angiography
MR angiography has no ionizing radiation and no iodine-containing contrast agents, and can sometimes replace conventional CT angiography. For CAFs, however, the biggest drawback of MR angiography was its limited effectiveness in showing the distal and extracardiac structures of fistula vessels (14, 15).
CTA
Compared with other imaging methods, CTA, especially dual-source CT, has the advantages of shorter examination time, higher temporal, spatial, density resolution, and it can better show origin of CAFs, course of fistula vessels, and drainage sites when combined with post-processing such as later volume rendering (VR), multiple Planar reorganization (MPR), curved surface reorganization (CPR). CTA is especially important in diagnosing CTAs with multiple fistula vessels, complex drainage sites, and small orificium fistulae (such as coronary-pulmonary fistula). This information is very important for surgeons and intervention doctors before surgery. These advantages of CTA also explained why the number of occasionally diagnosed CAFs increased with the popularity of CT (16–20). The obvious disadvantage of CT was its radiation exposure, which can be reduced by low-dose technologies such as prospective ECG gating and iterative reconstruction. Research results showed that the combination of tube voltage reduction and dose reduction technology can greatly reduce the radiation dose to 0.1 mSv (21, 22).
3. DSCT findings of CAF
The CT images of CAFs can be categorized mainly into two types. The first type was CAFs with a single origin, a single fistula vessel, and a single orificium fistulae. The affected coronary arteries dilated to varying degrees (typically the expansion and thickening show a "sausage-like" change), took a tortuous course and finally connected directly with the heart cavity or chest blood vessels. CT can clearly show the position and size of the orificium fistulae, and sometimes the contrast agent "jet sign" can be directly observed at the orificium fistulae (Fig. 1, Fig. 4). The second type was CAFs with multiple small or tortuous dilated malformed vessels from the left coronary artery, or right coronary artery or bilateral coronary artery. These vessels can communicate each other forming a reticular or diffuse disordered vessel cluster, and finally entangle and fistula into the main pulmonary artery or heart cavity and etc(Fig. 2). CAFs of this type were most common found in coronary-aortic fistulae. Orificium fistulae of this type was often very small, not often directly visible on DSCT, but in some cases the invisibility of orificium fistulae did not affect the qualitative diagnosis (23), and in some occasions the temporary differences in contrast agent concentration on both sides of orificium fistulae can indirectly indicate the presence of orificium fistulae (Fig. 3).
CAF vessels can originate from any of the three branches of the coronary artery, and the drainage site was more important than the origin in the clinic. We observed that among 62 cases of CAFs, the singly-originated left coronary artery accounted for the highest proportion with a total of 31 cases (50%), followed by 21 cases (34%) originating simultaneously from the left coronary artery and the right coronary artery, singly-originated right coronary arteries totaling 9 cases (15%), and the pulmonary artery has the largest number of drainage site (40/62), consistent with 66 CASs reported by Yang Shan (24), followed by right atrium (10/62) and right ventricle (7 / 62), which validated the view that the pulmonary artery with lower pressure and right heart system are the most common fistula access points (25–28). We observed that the most common type of coronary fistula was the coronary artery-pulmonary fistula, totaling 40 cases (65%), which contradicted the literature (3, 23, 29), though, we have obtained support from a relatively recent literature (30–32). We also noted that the left anterior descending coronary artery-aortic pulmonary fistula was the most common Coronary pulmonary-artery fistula (18/40), followed by the right corus branch-aortic fistula and left anterior descending coronary artery-aortic fistula (14/40). We hold that this is because on the one hand, previous CAG is an invasive test, some patients with asymptomatic coronary fistula will not be examined, and the dilution of CAG contrast agent may cause poor display of some small fistulae, so that some coronary artery-pulmonary fistulae are missed; on the other hand, with the popularization of CT, the detection rate of asymptomatic patients with Coronary pulmonary-artery fistula was getting higher and higher.
4.Management of CAFs
The clinical management strategy for patients with CAFs depended on the size and anatomy of the fistula vessel, presence or absence of symptoms, age of the patient, and presence of other relevant cardiovascular abnormalities (33). Asymptomatic CAFs with small orificium fistulae were usually treated with antiplatelet therapy and antibiotics, monitoring of complications and no need of intervention (2). According to the guidelines recommendation of American College of Cardiology and American Heart Association (34), treatment interventions can be considered in the following situations. First, large CAFs regardless of presence or absence of symptoms; Second, small to moderate-size fistulas (CAFs) with presence of symptoms. These symptoms included myocardial ischemia, arrhythmia, ventricular dysfunction, and endarteritis. Treatment options included surgical ligation and percutaneous transcatheter closure. Surgical ligation was generally recommended for the treatment of patients with large symptomatic CAFs, multiple communications, tortuous communications with aneurysmal arteries, and other heart diseases that require surgery treatment (13, 28). Percutaneous catheterization was a non-invasive treatment with a lower incidence of complications than surgical treatment (35). It was particularly suitable for patients who have surgery-benefited fistula anatomic location, single narrow communication, proximal end-originated fistula vessel, no multiple fistula vessels or large collateral vessels, or (and) no other associated heart disease (23). Many studies have shown that transcatheter closure was feasible, safe and effective, the results of which were almost the same as the surgical results in these cases (36, 37).
The shortcomings of this study were as follows. First the research sample was small, secondly this research was a single-center experience, so it was difficult to estimate the actual prevalence of CAFs in the general population and thirdly the design of this research was retrospective, thus some cases lacked clinical and CAG correlation.